THE SOURCES OF THE NITROGEN OF VEGETATION, ETC. 621 
the Graminacee and the Leguminose, which, in some points of view, appear to differ 
so widely in their demands upon combined Nitrogen provided within the soil. 
4, To take such as had already been experimented upon, with such conflicting results, 
by M. Boussineauut and M. G. Vitis. 
We shall first consider the results obtained with plants grown without any other 
supply of combined Nitrogen than that contained in the seed sown. 
I.—Experiments in which the plants had no other supply of combined Nitrogen than 
that contained in the seed sown. 
The following Table (XII.) gives, at one view, a summary of the numerical results 
obtained under this head; see also figs. 1-6, Plate XV., which are reduced from careful 
drawings taken of six out of the nine Graminacez experimented upon, and illustrate. 
_ the character and extent of growth attained under the conditions in question. 
After the full discussion in the foregoing pages of the circumstances under which the. 
results recorded in the Table just given were obtained, but little need be said in pointing 
out their bearings upon the question at issue. The column showing the gain or loss in 
each experiment speaks for itself. In judging of the results of the experiments of 
1857, the remarks made in discussing ,the results of Table XIV. (p. 532), with regard 
to the slates used as lute-vessels in that year, must be taken into consideration. The 
source of error referred to being obviated in the experiments of 1858, the results of 1857 
acquire a greater value, as confirming those of the latter year, than, standing alone, they 
would possess. 
The difference between the results obtained with soil and with pumice as matrix, in 
1857, are not such as to lead us to attach any importance to them, or to attribute them 
in any way to the difference of matrix in question. The two experiments may there- 
fore simply be considered as duplicates. Indeed, the character of the results in the one 
experiment with Wheat, and in the two with Barley, in 1857, was so similar, that the 
three experiments may be considered as triplicates. 
Graminaceous Plants. 
It will be observed that the largest gain of Nitrogen in the three experiments with 
Graminacee in 1857 was 0:0026 gramme. Keeping in view the probable source of 
error due to the use of slates in that year, and the difference of result in 1858 when 
slates were not employed, and, again, considering the fact that so small an amount of 
Nitrogen had to be determined in such a large amount of soil (0-003 gramme or less of 
Nitrogen in about 1500 grammes of soil), it seems indeed more than questionable whether 
the gain should not be attributed to the errors of experiment or analysis alluded to. In 
fact, we can but conclude that, under the circumstances of growth of the Graminaceous 
plants to which Table XII. relates, there has been no assimilation of free Nitrogen. 
It should also be noticed that, even when a gain of Nitrogen in the total products is 
observed, there is, in no case, more Nitrogen in the plant itself than in the original 
4B2 
